Multi-Agent Actor-Critic with Harmonic Annealing Pruning for Dynamic Spectrum Access Systems
This work addresses the problem of computational efficiency in multi-agent systems for dynamic spectrum access, offering an incremental improvement with a novel pruning scheduler.
The paper tackles the challenge of deploying deep learning models on resource-constrained edge devices for Dynamic Spectrum Access by proposing a sparse recurrent multi-agent reinforcement learning framework with harmonic annealing pruning, achieving superior policies that outperform conventional and state-of-the-art methods.
Multi-Agent Deep Reinforcement Learning (MADRL) has emerged as a powerful tool for optimizing decentralized decision-making systems in complex settings, such as Dynamic Spectrum Access (DSA). However, deploying deep learning models on resource-constrained edge devices remains challenging due to their high computational cost. To address this challenge, in this paper, we present a novel sparse recurrent MARL framework integrating gradual neural network pruning into the independent actor global critic paradigm. Additionally, we introduce a harmonic annealing sparsity scheduler, which achieves comparable, and in certain cases superior, performance to standard linear and polynomial pruning schedulers at large sparsities. Our experimental investigation demonstrates that the proposed DSA framework can discover superior policies, under diverse training conditions, outperforming conventional DSA, MADRL baselines, and state-of-the-art pruning techniques.